Cardiovascular diseases are common in avian species.1–5 Among these diseases, atherosclerosis is highly prevalent in aging parrots such as African grey parrots (Psittacus erithacus) and Amazon parrots (Amazona spp).2,6–10 Psittacine atherosclerotic lesions are caused by the progressive accumulation of inflammatory cells, cholesterol, fat, and cellular debris within the intima and on the luminal side of the tunica media of arteries, which ultimately results in luminal stenosis in major arteries.9,11 The antemortem diagnosis of atherosclerosis in birds is often difficult and represents a major challenge in avian medicine. Therefore, improvements in the available diagnostic methods, such as imaging techniques, are needed to facilitate the early detection of atherosclerosis in birds.12–14
Computed tomographic angiography is a potentially valuable diagnostic tool for detecting evidence of atherosclerotic lesions in birds.12 On the basis of data obtained from CTA with administration of contrast medium by use of bolus infusion or multiphase injection protocols, reference ranges for the diameters of major arteries in parrots have been published.13,14 A dual-head power injector has the potential to allow development of various new contrast medium injection protocols and to provide associated benefits such as ensuring vascular patency and enabling contrast-medium sparing and postinjection saline solution flushing, when compared with single-head injection methods. Various multiphase contrast medium injection protocols are included in guidelines for CTA in humans.15–17 However, current developments of CTA for use in birds do not include a guideline for the infusion protocol.
Compared with the test bolus method, the bolus tracking technique allows more efficient use of contrast medium with only a single injection of the full dose required for diagnostic purposes, and it is associated with shorter examination times and periods of general anesthesia.14 With the bolus tracking technique, data regarding time-enhancement variables are acquired and these illustrate the distribution of contrast medium in the body, which is frequently used to study the impact of different CTA protocols.18 The objective of the study reported here was to evaluate 3 CMI protocols for CTA and measurement of major artery diameters in African grey parrots.
Materials and Methods
Animals
Five male and 4 female 1- to 2-year-old African grey parrots with a mean body weight of 440 g (range, 345 to 495 g) were obtained from local pet shops. The birds were free of obvious clinical signs of cardiac disease, such as lethargy, dyspnea, coughing, or exercise intolerance. The birds were housed in separate cages (40 × 30 × 60 cm) in a well-ventilated room (4 × 5 × 5 m) at 25° to 30°C in the National Taiwan University Veterinary Hospital. The birds were fed a commercial diet twice daily and had clean drinking water provided ad libitum. Fruit and cuttlefish bone were supplied regularly, and direct sunlight was made available. The study protocol was approved by the National Taiwan University Institutional Animal Care and Use Committee.
Baseline assessment
A physical examination and auscultation of the lungs and heart were performed to ensure that each bird was healthy with clear lung sounds, regular heart rhythm, and no heart murmur. The following previously described diagnostic procedures19–22 were performed: blood analysis, including blood biochemical analysis,a protein electrophoresis,b and CBC; ventrodorsal and lateral whole-body radiographyc; 5-lead ECGd with 4 leads clamped on hypodermic needles positioned at the base of the 4 limbs and the cardiac lead clamped on a hypodermic needle positioned in the pectoral muscle just above the heart; and ventral midline approach echocardiography.e All of the procedures were performed while the birds were anesthetizedf; maintenance of anesthesia was provided by use of a veterinary anesthesia delivery system.g All of the measurements were compared with the published reference ranges for P erithacus.21–23 Various PCR analyses were used to detect polyomavirus (with samples from feathers, blood, and cloacal swabs), Chlamydia psittaci (with samples from conjunctival, choanal and cloacal swabs), and bornavirus (with samples from feathers, blood, and cloacal swabs) to rule out infectious diseases associated with cardiac disease.10,24,25 All of the birds were deemed healthy on the basis of the results of these examinations.
CTA and echocardiography
Computed tomographic angiography with each of 3 CMI protocols was performed in the order of protocols 1 through 3; the order assignment of the 9 parrots within a protocol assessment was randomized. The experiments for each protocol were completed for all birds before the next protocol was evaluated. However, if artifacts were present on the time-enhancement curve and contributed to poor imaging quality (as determined by the radiologist; Figure 1) for any bird, a second scan performed with that specific protocol was scheduled after all 3 protocols had been completed. If the second scan failed to produce a high-quality image, then no further attempts were allowed. The washout period between experiments, between the last experiment and any second scan, and between second scans (if there was more than 1) for all birds was at least 1 month.
Representative artifacts that appeared on the time-enhancement curves obtained by use of a bolus tracking technique during CTA of African grey parrots (Psittacus erithacus). The birds underwent each of 3 CMI protocols on separate occasions to evaluate the protocols for measurement of major artery diameters. A—Illustration of heterogeneity of contrast enhancement. The attenuation value fluctuated, with a marked increase as time progressed. This artifact was caused by uneven distribution of contrast medium in the bloodstream. B—Illustration of motion artifact. The preset ROI drifted away from the aorta and no enhancement was detected, leaving a blank image.
Citation: American Journal of Veterinary Research 79, 1; 10.2460/ajvr.79.1.42
For each protocol, each parrot was anesthetized by the same exotic animal anesthesiologist (CLC). Each parrot was placed in a plastic chamber, and induction of anesthesia was started with 5% isofluranef in oxygen at a flow rate of 2 L/min. The bird was subsequently intubated with an uncuffed No. 2 or No. 3 endotracheal tube,h and a veterinary anesthesia delivery systemg was used to maintain anesthesia with 2% to 3% isoflurane (adjusted depending on the depth of anesthesia of the bird). The machine was set to provide a flow rate of 0.8 L/min, breathing rate of 25 times/min, and peak inspiratory pressure of 7 to 8 mm Hg. Supplemental heat was provided by a heating lamp. A 24-gauge catheteri was placed in the left basilic vein of each parrot.
The dual-head injectorj was equipped with 2 independent and electrically connected syringes with Y-tubing and a single check valve. One of the syringes was front-loaded with contrast medium and the other was front-loaded with saline (0.9% NaCl) solution. During the CTA procedure, each parrot was placed in dorsal recumbency on an x-ray machine body cradle. The examinations were performed with a 16-detector row CT scanner.k Preliminary whole-body (scout) scans were first obtained in the craniocaudal direction, yielding the longitudinal field of view of the helical scan. On the basis of the scout images, helical scans without contrast medium administration were then performed from the last cervical vertebra to the hip joint to locate the ascending aorta with scanning settings as follow: section collimation, 1 mm; section width, 3 mm; scanning speed, 0.5 s/rotation; helical pitch, 1.0; table feed, 10 mm/s; peak electric potential, 120 kVp; and electrical current, 50 mA. During the scans obtained with protocols 1, 2, and 3, the same scanning settings were used, and a dynamic scan was performed simultaneously with a circular ROI in the middle portion of the ascending aorta (location determined by reference to unenhanced images) to monitor the attenuation values in time-enhancement curves. Image reconstruction with 1.0-mm thicknesses was performed by use of the scanner's workstation and commercial software.
The rate of IV injections of contrast medium and saline (0.9% NaCl) solution was 0.3 mL/s. Synchronization between administration of contrast medium and data acquisition was achieved with a realtime bolus-tracking method, and the scan was triggered manually once peak enhancement was reached (Figure 2). Three CMI protocols were evaluated. Protocol 1 involved a 3-phase CMI procedure. After flushing the IV catheter with 2.8 mL of saline solution to ensure patency, 2 mL of iopamidoll (a nonionic contrast agent) was infused, which was followed by a post-CMI flush with 0.2 mL of saline solution. The total infused volume for protocol 1 was 5 mL. Protocol 2 involved a 2-phase CMI procedure. After 2 mL of iopamidol was infused, a post-CMI flush with 0.4 mL of saline solution was administered. The total infused volume for protocol 2 was 2.4 mL. Protocol 3 involved another 2-phase CMI procedure without a post-CMI saline solution flush. After flushing the IV catheter with 2.8 mL of saline solution, 2 mL of iopamidol was infused. The total infused volume for protocol 3 was 4.8 mL.
Representative image of a successful time-enhancement curve obtained from an African grey parrot by use of a bolus tracking technique during CTA involving a 2-phase CMI procedure (protocol 2). After 2 mL of iopamidol was infused, a post-CMI flush with 0.4 mL of saline (0.9% NaCl) solution was administered. The total infused volume for protocol 2 was 2.4 mL. Notice that there was still initial heterogeneity of contrast enhancement, which causes a decrease in the attenuation value; however, peak enhancement was attained within 5 seconds and the scan was triggered manually thereafter.
Citation: American Journal of Veterinary Research 79, 1; 10.2460/ajvr.79.1.42
Once the protocol experiment was initiated for each bird, maintenance of anesthesia was provided by inhalation of 5% isoflurane in oxygen and anesthesia settings were adjusted as follows: flow rate, 0.6 mL/min; breathing rate, 2 times/min; and peak inspiratory pressure, 5 mm Hg. These adjustments were made to diminish motion artifacts and enable data acquisition. The tissue oxygenation and heart rate were monitored with a pulse oximeter. The IV catheter was removed after the scan completed. A final physical examination, which included auscultation of the lungs and heart, was performed after the birds recovered from anesthesia. Each bird's overall condition including appetite, responsiveness, and behavior was monitored for 3 days after the experiments.
ROI measurements
A window level of 40 HU and a window width of 400 HU were used for measurement of the diameters of 6 contrast medium–enhanced major arteries, which included the left and right brachiocephalic trunks, ascending aorta, left and right pulmonary arteries, and abdominal aorta. The circular ROI was obtained by encircling the lumen of the vessel with the circular measurement tool of the picture archiving and communication imaging system.m The blurred edges were excluded to reduce inaccuracy by avoiding possible blooming and penumbra artifacts caused by the scanning settings or the bird's breathing. The cross-sectional area of the lumen (A) was calculated automatically by the imaging softwarem (graphics device interface mode with software bilinear stretch), and the arterial diameter (D) was calculated manually as D = 2(A/π)1/2.
The left and right brachiocephalic trunks were measured where they crossed the cranial thoracic cavity laterally; however, the measurement may not have been taken at the same level in each vessel because there is a bulging section at the end of the right brachiocephalic trunk and the measurement was obtained at a slightly more proximal location (Figure 3). The ascending aorta was measured at the middle of its ascending course. The abdominal aorta was measured proximal to the bifurcation of the mesenteric artery. The left and right pulmonary arteries were each measured at the middle of their ascending course to the lungs. Cross-sectional images were magnified at 200% to improve the accuracy of the measurements. Two observers participated in ROI measurements. Observer 1 obtained all measurements twice on different days to evaluate intra-observer agreement of the measurement techniques. Additionally, observer 2 obtained measurements once to evaluate interobserver agreement.
Representative CTA images obtained at the level of the brachiocephalic trunks (A and B [arrow]), ascending aorta (C [arrow]), left pulmonary artery (D [arrow]), right pulmonary artery (E [arrow]), and abdominal aorta (F [arrow]) in an African grey parrot. The parrot had undergone a 2-phase CMI procedure (protocol 2). In panel A, notice that the bulging section at the end of the right brachiocephalic trunk may suggest a change in the general direction of the artery at this level. As a result, the circular ROI used to measure the diameter of the vessel was placed on the CT slice preceding the slice that would have been used. Thus, the diameters of the right and left brachiocephalic trunks were measured at slightly different levels of the vessel.
Citation: American Journal of Veterinary Research 79, 1; 10.2460/ajvr.79.1.42
Time-enhancement variables
On the basis of previously obtained data,22 the mean ± SD diameter of the ascending aorta of African grey parrots (3.58 ± 0.44 mm) is comparable to the diameter of a human coronary artery (3.4 to 4.2 mm).26 The optimal attenuation level for human coronary arteries is suggested be at least 250 HU, with targets of 300 and 350 HU.27 Because of their small body size and fast heart rate and circulation of contrast medium in the vascular system, greater enhancement may be needed for adequate visualization of the ascending aorta in parrots13; thus, the lower range was set at 330 HU. The time required to attain 330 HU, indicating initiation of the optimal enhancement period, was designated as T1. The aortic attenuation measured at the peak of the time-enhancement curves was designated as Hmax, and the time from start of contrast medium injection to Hmax was designated as Tmax.
Statistical analysis
All of the statistical analyses were performed with computer software.n A Kolmogorov-Smirnov test was used to evaluate the distribution of the data.28 For data that were not normally distributed, the median and range are reported. For data that were normally distributed, the mean ± SD are reported. Differences among the CTA-derived measurements of vessel diameters obtained with the 3 CMI protocols were evaluated with repeated-measures general linear models. Intra-observer agreement (reproducibility) between 2 consecutive measurements was evaluated by calculating the Spearman correlation coefficient (ρ) for each artery. Reproducibility was considered high when the coefficient was ≥ 0.76, medium when the coefficient was 0.40 to 0.75, and low when the coefficient was ≤ 0.39. Agreement between the 2 observers was evaluated by calculating the ICC. The 2-way random, single-measure reliability ICC for absolute agreement and 95% confidence intervals were determined for each artery. Interobserver agreement was considered high when the coefficient was ≥ 0.76, medium when the coefficient was 0.40 to 0.75, and low when the coefficient was ≤ 0.39. Values of P < 0.05 were considered significant for all analyses.13
Measurements of T1, Tmax, and Hmax were used to evaluate the time-enhancement curve; the differences among the protocols were tested by ANOVA, as was the effect of body weight when the parrots were grouped by weight as follows: weight group 1, 470 to 505 g (n = 3); weight group 2, 400 to 450 g (4); and weight group 3, < 350 g (1). The effect of sex was tested with a t test. The level of significance for all of these analyses was set at a value of P < 0.05.
Results
Baseline assessment
The results of the blood analysis were within the reference ranges for all 9 birds.25 The 2-view radiographic and ECG findings were unremarkable.20–22 No infectious disease was detected in any bird. Some baseline echocardiographic measurements differed slightly from published reference ranges, including lower systolic left ventricular length and greater systolic interventricular septum thickness.15 However, the importance of slight differences in avian echocardiographic measurements is still undetermined, and no abnormalities in cardiac structure or function were detected. Therefore, all birds were deemed healthy and free of detectable cardiovascular disease.29
CTA measurements
The CTA scans and CMI procedures were successfully completed for 8 of the 9 birds undergoing protocol 1 (14 scans) and for all 9 birds undergoing protocol 2 (12 scans) and protocol 3 (18 scans). Among a total of 44 scans, heterogeneity of contrast enhancement resulted in poor image quality in 5, 3, and 9 of the scans obtained for protocols 1, 2, and 3, respectively. Motion artifact resulted in poor image quality in 0, 1, and 1 of the scans obtained for protocols 1, 2, and 3, respectively. No extravasation of contrast medium or saline solution was observed in any image.
CTA measurements of vessel diameter obtained with the 3 CMI protocols
The arterial diameters and attenuation values determined by means of CTA were summarized (Table 1). The mean diameters of the right brachiocephalic trunk and ascending aorta determined from images obtained with protocol 2 were significantly larger than those determined from images obtained with protocol 1. The mean diameters of the ascending aorta, left and right brachiocephalic trunks, left pulmonary artery, and abdominal aorta determined from images obtained with protocol 2 were significantly larger than those determined from images obtained with protocol 3. The diameters of the ascending aorta, left pulmonary artery, and abdominal aorta determined from images obtained with protocol 1 were significantly larger than those determined from images obtained with protocol 3 (Table 2).
Attenuation values and arterial diameters determined by use of CTA with 3 CMI protocols in 9 African grey parrots (Psittacus erithacus).
| Protocol 1 | Protocol 2 | Protocol 3 | ||||
|---|---|---|---|---|---|---|
| Artery | Attenuation value (HU) | Diameter (mm) | Attenuation value (HU) | Diameter (mm) | Attenuation value (HU) | Diameter (mm) |
| Right brachiocephalic trunk | 421.84 ± 126.12 | 3.04 ± 0.53 | 514.34 ± 73.57 | 3.58 ± 0.48*† | 242.74 ± 92.67 | 2.53 ± 0.63 |
| Left brachiocephalic trunk | 420.39 ± 129.68 | 2.69 ± 0.53 | 483.53 ± 75.10 | 3.20 ± 0.60† | 159.5 (84.1–392.3) | 2.45 ± 0.60 |
| Ascending aorta | 338.24 ± 113.38 | 2.93 ± 0.67‡ | 507.58 ± 64.36 | 3.58 ± 0.76*† | 205.61 ± 40 | 2.53 ± 0.54 |
| Right pulmonary artery | 383.08 ± 77.09 | 1.93 ± 0.26 | 426.31 ± 47.50 | 0.22 (0.16–0.34) | 204.81 ± 115.40 | 0.12 (0.12–0.22) |
| Left pulmonary artery | 406.96 ± 151.84 | 1.91 ± 0.36‡ | 432.49 ± 78.52 | 0.22 (0.16–0.32)† | 157. (126.5–356.3) | 0.12 (0.12–0.22) |
| Abdominal aorta | 329.26 ± 76.30 | 0.16 (0.12–0.2)‡ | 367.31 ± 30.02 | 0.22 (0.16–0.34)† | 198.18 ± 72.49 | 0.12 (0.12–0.16) |
Data that were normally distributed are reported as mean ± SD; data that were not normally distributed are reported as median and range.
Value obtained with protocol 2 is significantly (P < 0.05) different from that obtained with protocol 1.
Value obtained with protocol 2 is significantly (P < 0.05) different from that obtained with protocol 3.
Value obtained with protocol 1 is significantly (P < 0.05) different from that obtained with protocol 3. Attenuation values of each artery were not statistically compared because comparison of Hmax in time-enhancement curves and the signal-noise ratio will reflect the quality of contrast enhancement more clearly.
For the 9 birds, CTA with each of 3 CMI protocols was performed in the order of protocols 1 through 3. The experiments for each protocol were completed for all birds before the next protocol was evaluated. However, if artifacts were present on the time-enhancement curve and contributed to poor imaging quality for any bird, a second scan performed with that specific protocol was scheduled after all 3 protocols had been completed. If the second scan failed to produce a high-quality image, then no further attempts were allowed. The washout period between experiments and any additional scans for all birds was at least 1 month. For each experiment, the birds were anesthetized. The rate of IV injection of contrast medium and saline (0.9% NaCl) solution was 0.3 mL/s. Protocol 1 (total infused volume, 5 mL) involved a 3-phase CMI procedure. After flushing the IV catheter with 2.8 mL of saline solution to ensure patency, 2 mL of iopamidol was infused, which was followed with a post-CMI flush with 0.2 mL of saline solution. Protocol 2 (total infused volume, 2.4 mL) involved a 2-phase CMI procedure. After 2 mL of iopamidol was infused, a post-CMI flush with 0.4 mL of saline solution was administered. Protocol 3 (total infused volume, 4.8 mL) involved another 2-phase CMI procedure without a post-CMI saline solution flush. After flushing the IV catheter with 2.8 mL of saline solution, 2 mL of iopamidol was infused. A circular ROI was obtained by encircling the lumen of the vessel with the circular measurement tool of a picture archiving and communication imaging system. The cross-sectional area of the lumen (A) was calculated automatically by imaging software, and the arterial diameter (D) was calculated manually as D = 2(A/π)1/2. The left and right brachiocephalic trunks were measured where they crossed the cranial thoracic cavity laterally. The ascending aorta was measured at the middle of its ascending course. The abdominal aorta was measured proximal to the bifurcation of the mesenteric artery. The left and right pulmonary arteries were each measured at the middle of the ascending course to the lungs. The CTA scans and CMI procedures were successfully completed for all 9 birds undergoing protocols 2 and 3, and for 8 of the 9 birds undergoing protocol 1.
Differences (mm) in measurements of diameters of 6 major arteries determined by CTA with each of 3 CMI protocols for 9 African grey parrots.
| Right brachiocephalic trunk | Left brachiocephalic trunk | Ascending aorta | Right pulmonary artery | Left pulmonary artery | Abdominal aorta | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Protocol comparison | Difference | P value | Difference | P value | Difference | P value | Difference | P value | Difference | P value | Difference | P value |
| 1 vs 2 | –0.63* | 0.016 | –0.55 | 0.064 | –0.55* | 0.024 | 0.14 | 0.572 | 0.003 | 0.871 | –0.15 | 0.197 |
| 2 vs 3 | 0.10* | 0.002 | 0.75* | 0.002 | 0.08* | 0.002 | 0.04 | 0.872 | 0.53* | 0.034 | 0.45* | < 0.05 |
| 3 vs 1 | –0.37 | 0.069 | –0.20 | 0.342 | –0.25* | 0.028 | –0.18 | 0.474 | –0.55* | 0.041 | –0.30* | 0.048 |
For a given artery, the difference in diameter determined during use of the 2 specified protocols was significant (P < 0.05). See Table 1 for key
Intra-observer and interobserver agreements were high for larger arteries, such as the ascending aorta and brachiocephalic trunks, for images obtained with all 3 protocols; the Spearman correlation coefficient and ICC were ≥ 0.830 with the exception of medium intra-observer agreement for the ascending aorta in birds undergoing protocol 3 (coefficient, 0.739; Table 3). The intra-observer and interobserver agreement for the pulmonary arteries was medium to high for images obtained with all 3 protocols. However, for the abdominal aorta, the interobserver agreement in images obtained with protocol 3 was low, with an ICC of 0.148.
Intra-observer agreement (reproducibility) for 1 observer and interobserver agreement between 2 observers for CTA-derived measurements of diameters of 6 major arteries determined for 9 African grey parrots that underwent each of 3 CMI protocols.
| Right brachiocephalic trunk | Left brachiocephalic trunk | Ascending aorta | Right pulmonary artery | Left pulmonary artery | Abdominal aorta | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Protocol | Coefficient | 95% CI | Corr | 95% CI | Corr | 95% CI | Corr | 95% CI | Corr | 95% CI | Corr | 95% CI | Corr |
| 1 | SCC | — | 0.897 | — | 0.870 | — | 0.901 | — | 0.766 | — | 0.822 | — | 0.887 |
| ICC | 0.67 to 0.98 | 0.914 | 0.59 to 0.97 | 0.891 | 0.60 to 0.98 | 0.894 | –0.12 to 0.88 | 0.549 | 0.04 to 0.91 | 0.655 | 0.53 to 0.97 | 0.871 | |
| 2 | SCC | — | 0.886 | — | 0.956 | — | 0.983 | — | 0.703 | — | 0.782 | — | 0.316 |
| ICC | 0.62 to 0.98 | 0.899 | 0.86 to 0.99 | 0.967 | 0.79 to 0.99 | 0.948 | 0.06 to 0.91 | 0.664 | 0.57 to 0.97 | 0.883 | 0 to 0.90 | 0.632 | |
| 3 | SCC | — | 0.945 | — | 0.876 | — | 0.739 | — | 0.928 | — | 1 | — | 1 |
| ICC | 0.76 to 0.99 | 0.947 | 0.78 to 0.99 | 0.952 | 0.37 to 0.96 | 0.830 | 0.49 to 0.97 | 0.87 | 0.24 to 0.95 | 0.783 | –0.58 to 0.74 | 0.148 | |
Two observers participated in ROI measurements (a circular ROI was obtained by encircling the lumen of the vessel in the CTA image with the circular measurement tool of a picture archiving and communication imaging system). Observer 1 obtained all measurements twice on different days to evaluate intra-observer agreement of the measurement techniques. Additionally, observer 2 obtained measurements once to evaluate interobserver agreement.
— = Not applicable. CI = Confidence interval. Corr = Correlation. ICC = Two-way, random, single-measure reliability ICC for absolute agreement, comparing the 2 observers’ measurements. SCC = Spearman correlation coefficient for 2 measurements made by 1 individual.
Time-enhancement curves obtained with the 3 CMI protocols
Time-enhancement curves were obtained for the birds undergoing the 3 CMI protocols (Figures 4, 5, and 6). Of the 9 birds, only 3 and 2 had an aortic enhancement value > 330 HU when undergoing protocols 1 and 3, respectively; on the other hand, 8 of 9 birds had an aortic enhancement value > 330 HU when undergoing protocol 2. The mean values of T1 and Tmax were significantly (both P < 0.05) smaller for birds undergoing protocol 2 than for birds undergoing protocol 1 during a period > 10 seconds. Similar differences were also observed for birds undergoing protocol 2, compared with birds undergoing protocol 3 (both P < 0.05). On the other hand, mean Hmax for birds undergoing protocol 2 was significantly higher than that for birds undergoing protocol 3. Although there was no significant difference between the values for Hmax obtained for protocols 2 and 1 (P = 0.054), mean Hmax for birds undergoing protocol 2 was 108 HU larger than that obtained with protocol 1. The time-enhancement data indicated that contrast enhancement obtained with protocol 2 was more uniform and efficient and had less heterogeneity, compared with data for the other 2 protocols.
Time-enhancement curves obtained from 8 of 9 African grey parrots by use of a bolus tracking technique during CTA involving a 3-phase CMI procedure (protocol 1). After flushing the IV catheter with 2.8 mL of saline solution to ensure patency, 2 mL of iopamidol was infused, which was followed with a post-CMI flush with 0.2 mL of saline solution. Each time-enhancement curve was obtained from a different bird. One of the curves is not included because the quality of enhancement was considered a failure. Only 3 of the 9 birds had an aortic enhancement value > 330 HU, and overall mean T1 (time to attain 330 HU) and Tmax were significantly longer than those obtained when birds underwent protocol 2. See Figure 2 for remainder of key.
Citation: American Journal of Veterinary Research 79, 1; 10.2460/ajvr.79.1.42
Time-enhancement curves obtained from 9 African grey parrots by use of a bolus tracking technique during CTA involving a 2-phase CMI procedure (protocol 2). Each time-enhancement curve was obtained from a different bird, and individual curves are displayed following the same order and scale as those in Figure 4. Protocol 2 resulted in a markedly uniform waveform, less heterogeneity of contrast enhancement, and a higher peak attenuation value, compared with the results of protocols 1 or 3. Protocol 3 was a 2-phase CMI procedure without a post-CMI saline solution flush; after flushing the IV catheter with 2.8 mL of saline solution, 2 mL of iopamidol was infused. See Figures 2 and 4 for remainder of key.
Citation: American Journal of Veterinary Research 79, 1; 10.2460/ajvr.79.1.42
Time-enhancement curves obtained from 9 African grey parrots by use of a bolus tracking technique during CTA involving a 2-phase CMI procedure (protocol 3). Each time-enhancement curve was obtained from a different bird, and curves are displayed following the same order and scale as those in Figure 4. Only 2 of the 9 birds had an aortic enhancement value > 330 HU, and overall mean T1 and Tmax were significantly longer than those obtained when birds underwent protocol 2. See Figures 2 and 4 for remainder of key.
Citation: American Journal of Veterinary Research 79, 1; 10.2460/ajvr.79.1.42
Of the 3 CMI protocols, only protocol 2 was associated with a marked success rate, if one considers an attenuation value > 330 HU as indication of satisfactory enhancement of major arteries. Therefore, we compared different patient factors within the results for protocol 2. There was no significant difference in mean Hmax (P = 0.502), Tmax (P = 0.652), or T1 (P = 0.644) among the 3 body weight groups. Also, there were no significant difference between males and females with regard to mean Hmax (P = 0.735), Tmax (P = 0.622), or T1 (P = 0.78 9).
During recovery following the protocol 1 experiments, 2 of the 9 birds developed mild bilateral lung crackling immediately after the experiment (detected during auscultation). No lung crackling was ausculted in birds after having undergone the protocol 2 and protocol 3 experiments. All birds were considered clinically normal with normal appetite, responsiveness, and behavior within 2 hours after recovery from anesthesia.
Discussion
In the study reported here, 3 CMI protocols for CTA in African grey parrots were proposed and comparisons of the apparent luminal arterial diameters and time-enhancement curves were made to determine the best infusion protocol for this species of bird. Given the significant differences in most of the arterial diameters among the 3 CMI protocols, determination of the appropriate protocol is warranted in this species. In the present study, protocol 2, which involved use of a dual-head power injector to provide an IV infusion of iopamidol (2 mL) followed by a post-CMI flush of saline solution of 0.4 mL, produced the best results, as indicated by the comparatively superior intra- and interobserver agreements, time-enhancement curve, artifact occurrence rate, and complication rate. Nevertheless, on the basis of the comparisons among protocols in the present study and on results of previous studies,13,14 further modification of protocol 2 is warranted.
To assess the minimum volume of post-CMI saline solution flush needed, it was critical to estimate the volume of space between the injector tube and the ascending aorta; the dead space volume can be approximated by multiplication of the time taken for the contrast medium to reach the ROI and the infusion rate.o On the basis of results of previous trials, increasing contrast enhancement at the ROI was evident at approximately 1 second after the start of the contrast medium injection; therefore, we assumed that the time taken for the contrast medium to reach the ROI was 1 second. Adjustment of the injection rate is recommended to reduce the variability of contrast enhancement caused by cardiac output, which is the main physiologic variable that controls contrast enhancement.18,30 In the present study, we used published data regarding the cardiac output and injection rate in different species as a guideline for the adjustment. The cardiac output of rats, pigeons, and rabbits is 80 mL/min, 245 mL/min, and 600 to 800 mL/min, respectively,31–33 and the infusion rate for rats and rabbits is 0.1 mL/s and 0.5 mL/s, respectively.34,35 Therefore, the infusion rate for the study parrots was set at 0.3 mL/s, a rate similar to that used in a previous study.13 Thus, the minimum volume of post-CMI saline solution flush was 0.3 mL/s multiplied by 1 second. In the present study, we set the volume of the post-CMI saline solution flush at 0.4 mL. The significantly higher Hmax, shorter T1, and shorter Tmax in birds undergoing protocol 2 confirmed the contrast-sparing potential of the post-CMI saline solution flush and proved that this flush volume should be sufficient to fill the dead space before the ROI.
Heterogeneity of contrast enhancement was a major contributor to the artifacts detected during the present study. When encountered, the cross-sectional image of the ascending aorta on the dynamic scan would appear inconsistently and fuzzily enhanced, making the edge of the arterial lumen blurry and the attenuation values fluctuate. The heterogeneity of contrast enhancement was caused by the extremely rapid travel of the contrast medium in the circulation owing to the high cardiac output in birds. Although the post-CMI saline solution flush used in protocol 2 efficiently decreased the amount of heterogeneity of the contrast enhancement, there was still a mild degree of concomitant hypo- and hyperenhancement within the arterial lumen. To overcome this problem, an injection rate > 0.3 mL/s should be tested in the future.
Compared with previously described protocols,13 we used a higher dose of contrast medium (740 mg of iodine/bird) than those used in Amazon parrots (256.8 to 300 mg of iodine/bird). The suggested dose for avian angiography ranges from 300 to 1,520 mg of iodine/kg.13,36,37 Given the difference in size between African grey parrots and Amazon parrots, we intended to better define the iodine dose range for African grey parrots and selected an iodine dose that was greater than that suggested for Amazon parrots but less than the maximum suggested dose among avian species. In birds undergoing protocol 2 in the present study, higher mean attenuation values and smaller SDs in all great arteries, compared with results of another study in Amazon parrots,13 indicated better image quality with respect to the signal-to-noise ratio (ie, mean attenuation [HU]/SD [HU]) for the image quality assessment system (Table 4), and suggested that reductions of the amount of radiation and contrast medium dose may be possible.38,39 In addition to variations in the doses of contrast medium and radiation, there were differences in injection duration and scanning time between the present study and the study in Amazon parrots.13 In protocol 2, the total injection time was extended from 3 seconds previously described in the Amazon parrot study1,3 to 8 seconds in the present study by inclusion of a post-CMI saline solution flush; this additional flush allowed contrast medium in the dead space to be propelled further into the ROI and delayed scanning time after the CMI from 3 seconds to 4 to 8 seconds. In addition, this additional flush extended the optimal scanning time during which the attenuation value could be maintained at a high level for a longer period, thereby avoiding interference caused by heterogeneity of contrast enhancement.
Signal-to-noise ratio determined by use of CTA and CMI protocol 2 for 6 major arteries in 9 African grey parrots.
| Right brachiocephalic trunk | Left brachiocephalic trunk | Ascending aorta | Right pulmonary artery | Left pulmonary artery | Abdominal aorta | |
|---|---|---|---|---|---|---|
| Mean attenuation (HU) | 514.34 | 483.53 | 507.58 | 426.31 | 432.49 | 367.31 |
| Attenuation SD (HU) | 73.56 | 75.10 | 64.36 | 47.50 | 78.52 | 30.02 |
| Signal-to-noise ratio | 6.99 | 6.43 | 7.89 | 8.97 | 5.51 | 12.24 |
Signal-to-noise ratio was calculated as mean attenuation (HU) divided by the attenuation SD (HU). See Table 1 for key.
The levels of intra-observer and interobserver agreements in the present study were not as high as those previously reported,13 especially for smaller arteries. In the present study, the software could only measure the circled area of the arteries to 2 decimal places, and it was not sensitive enough to identify minor differences in the ROI in smaller arteries. Therefore, minor differences in the measurements could be rounded off and result in more imprecise readings for the circled areas, making the calculated diameters variable and reducing the level of agreement. The small sample size of the present study likely exaggerated the situation. Finally, the ROI location (Figure 3) of each major artery was somewhat different from that previously described for Amazon parrots13; species differences may make the levels of agreement not comparable between the present study and the study in Amazon parrots.13
To reduce measurement variability during CTA, manipulation of the background condition could be valuable. Window and level settings significantly affect the accuracy of CTA in humans, and optimized settings have been proposed regarding the contrast attenuation coefficient of carotid artery.40 In parrots, arterial measurements are window dependent and a mediastinal or angiography window is recommended.13,14 Thus, tailoring the window and level settings for individual birds could also have affected the results and potentially have improved the intra- and interobserver agreements in the present study. On the other hand, the accuracy of diameter measurement was also affected by the convolution kernel. A study41 involving vascular models revealed that use of higher-density contrast medium and selection of the proper convolution kernel will improve accuracy of vessel diameter measurement. In that study,41 the soft convolution kernel was better than the standard convolution kernel for the measurement of small artery diameters. In the present study, the kernel used during data acquisition and diameter measurement was 17. Two limits of convolution kernel that we commonly use in clinical practice are 3 and 52, and contrast resolution tends to decrease as the limit increases. Further evaluation of the effect of different convolution kernels may be warranted, especially when measuring the internal diameter of the small arteries of psittacine species. Automated software is expected to decrease operator dependence and improve objectivity in arterial diameter measurement. However, such assumptions are valid only when the density of the intravascular contrast medium is adequate.41 On the other hand, manual measurement of the internal diameter of vessels in images is easy to standardize and is associated with excellent interobserver agreement.42 Various techniques and algorithms to evaluate different arteries with the highest accuracy in humans have been proposed.42 Regarding CTA in parrots, the usefulness of automated software may not currently outweigh that of manual measurement because of the lack of information regarding optimal scanning settings, but assessment of methods to improve diagnostic accuracy are certainly necessary.
In the present study, the use of a 3-phase CMI protocol was based on CTA procedures in humans and was selected because of the various aforementioned benefits.15–17 For comparison, protocols 2 and 3 were designed without the pre-CMI saline solution flush and post-CMI saline solution flush, respectively, to verify the effects of saline solution infusion on the images. A high total injection volume requires a prolonged infusion during which the injection rate is fixed. The total duration of the injections is the most important factor with regard to the timing of a CT scan because it directly affects the time to peak contrast enhancement in an organ or vessel of interest.16 In both protocols 1 and 3, the volume of pre-CMI saline solution flush (2.8 mL) was fixed by the Y-shaped connecting tube of the dual-head injector, and the total infused volume was at least twice that used in protocol 2. This large difference in total infusion volume relative to the body size of a parrot may influence the optimal timing for CT scanning by delaying peak contrast enhancement in the vessel of interest and leading to a failure to reach the desired level of enhancement.43 In addition, the pre-CMI saline solution flush used in protocols 1 and 3 may have diluted the enhancement effect of the contrast medium and increased the heterogeneity of contrast enhancement. Given that no extravasation was detected in all 44 scans obtained in the present study, the pre-CMI saline solution flush may not be needed.
The findings of the present study were subject to certain limitations. To record peak contrast enhancement, CT scans were started 2 seconds after peak enhancement was achieved, and the attenuation value may have been lower at this point. However, the delay was necessary because the time-enhancement curve was no longer available once the scanning had started. It was found, especially in protocol 2, that use of the post-CMI saline solution flush gained extra seconds of optimal scanning time and that the image quality was still good, as indicated by the signal-to-noise ratio. Nevertheless, a complete time-enhancement curve specific for each major artery is still needed for the development of optimal scanning settings.
In the present study, there were some factors that were potential limitations for a proposed CTA protocol in African grey parrots. We overcame most motion artifact by adjustment of the anesthetic protocol, but motion artifact attributable to the absence of an ECG-gated scan was still present. An ECG-gated scan has been shown to improve image quality and increase diagnostic accuracy in human neonates and infants.44,45 However, results of some studies also suggest that the principle of performing scans as fast as possible is adequate for obtaining good images, minimizing respiratory artifact, and reducing radiation exposure in neonates with coronary heart disease.46 In addition, owing to the limited number of studies of CTA in birds, the interpretation of the data obtained in the present study relied heavily on the findings of studies of humans and small mammals. Finally, the rate of the CMI and the amount of contrast medium administered as well as the size of the catheter used in the study birds could have interfered with the results.
By combining a dual-head injector and a bolus tracking technique in the present study, we were able to conclude that a 2-phase CMI protocol that included a post-CMI saline solution flush provided more reliable CTA-derived measurements of major thoracic and abdominal arteries in African grey parrots, compared with a 3-phase CMI protocol (comparable to protocols used in humans) and a 2-phase CMI protocol with no post-CMI saline solution flush. Given the study findings, further optimization of the infusion rate and appropriate volumes of contrast medium and saline solution to attain optimal contrast enhancement should be possible. Further study of CTA and CMI protocols in birds is warranted for this technique to become diagnostically useful in cases of avian atherosclerosis.
Acknowledgments
The authors thank Dr. Yong-Wei Hong and Da-Li Lu of Cardiospecial Veterinary Hospital, Taipei, Taiwan, for advice regarding echocardiography; Ai-Ting Lin of the National Taiwan University Veterinary Hospital, Taipei, Taiwan, for assistance in operating the CT scanner and troubleshooting the procedures; and Dr. Ya-Pei Chang and Dr. Kwo-Chen Hu for helpful discussions and the use of the CT scanner.
ABBREVIATIONS
| CMI | Contrast medium infusion |
| CTA | Computed tomographic angiography |
| Hmax | Aortic attenuation measured at the peak of the time-enhancement curve |
| HU | Hounsfield unit |
| ICC | Interclass correlation coefficient |
| ROI | Region of interest |
| Tmax | Time from start of contrast medium injection to Hmax |
Footnotes
VITROS 350 Chemistry System, Ortho Clinical Diagnostics, Johnson & Johnson, Melbourne, Australia.
SPIFE 3000 Gel Electrophoresis and Processing, Helena Laboratories, Beaumont, Tex.
Toshiba KXO-32s, Toshiba Medical Systems Corp, Tokyo, Japan.
AT-1 smartprint ECG unit, Schiller AG, Baar, Switzerland.
EnVisor C HD with S12 UltraBand Sector Xducer, Philips Electronics North America Corp, Andover, Mass.
Attane, Isoflurane, Panion & BF Biotech Inc, Taoyuan, Taiwan.
A.D.S 2000; ENGLER, Hialeah, Fla.
Mallinckrodt Oral/Nasal Cuffless Endotracheal Tube, Covidien Taiwan Ltd, Taipei, Taiwan.
Surflo Etfe IV catheter 24G × 3/4”, TERUMO (Philippines) Corp, Biñan, Laguna, Philippines.
OptiVantage Dual-Head CT Contrast Delivery System, Mallinckrodt Pharmaceuticals, Dublin, Ireland.
Activion 16, Toshiba Corp, Tokyo, Japan.
Iopamiro 370, Bracco s.p.a., Milan, Italy.
Uniweb Viewer, version 7.2.1524, EBM Technologies Inc, Taipei, Taiwan.
IBM SPSS Statistics for Windows, Version 21.0, IBM Corp, Armonk, NY.
Kuroki Y, Yamabe Y, Yoshida Y, et al. Estimation of ‘dead space’ volume using MDCT: assessing the minimal saline flush volume after a bolus injection of contrast medium for optimal arterial phase imaging in the dynamic study, in Proceedings. Euro Soc Radiol 2010; dx.doi.org/10.1594/ecr2010/C-1131.
References
1. Bavelaar FJ, Beynen AC. Atherosclerosis in parrots: a review. Vet Q 2004; 26: 50–60.
2. Krautwald-Junghanns ME, Braun S, Pees M, et al. Research on the anatomy and pathology of the psittacine heart. J Avian Med Surg 2004; 18: 2–11.
3. Oglesbee BL, Oglesbee MJ. Results of postmortem examination of psittacine birds with cardiac disease: 26 cases (1991–1995). J Am Vet Med Assoc 1998; 212: 1737–1742.
4. Schmidt RE, Hubbard GB, Fletcher KC. Systematic survey of lesions from animals in a zoologic collection. I. Central nervous system. J Zoo Anim Med 1986; 17: 8–11.
5. St. Leger J. Avian atherosclerosis. In: Fowler ME, Miller RE, eds. Zoo and wild animal medicine, current therapy. 6th ed. St Louis: Saunders, 2007;200–205.
6. Bohorquez F, Stout C. Aortic atherosclerosis in exotic avians. Exp Mol Pathol 1972; 46: 11–19.
7. Pilny AA. Retrospective of atherosclerosis in psittacine birds: clinical and histopathologic findings in 31 cases, in Proceedings. Assoc Avian Vet Conf 2004;349–351.
8. Garner MM, Raymond JT. A retrospective study of atherosclerosis in birds, in Proceedings. Assoc Avian Vet Conf 2003;59–66.
9. Fricke C, Schmidt V, Cramer K, et al. Characterization of atherosclerosis by histochemical and immunohistochemical methods in African grey parrots (Psittacus erithacus) and Amazon parrots (Amazona spp.). Avian Dis 2009; 53: 466–472.
10. Beaufrère H, Ammersbach M, Reavill DR, et al. Prevalence of and risk factors associated with atherosclerosis in psittacine birds. J Am Vet Med Assoc 2013; 242: 1696–1704.
11. Beaufrère H, Nevarez JG, Holder K, et al. Characterization and classification of psittacine atherosclerotic lesions by histopathology, digital image analysis, transmission and scanning electron microscopy. Avian Pathol 2011; 40: 531–544.
12. Beaufrère H, Pariaut R, Rodriguez D, et al. Avian vascular imaging: a review. J Avian Med Surg 2010; 24: 174–184.
13. Beaufrère H, Rodriguez D, Pariaut R, et al. Estimation of intrathoracic arterial diameter by means of computed tomographic angiography in Hispaniolan Amazon parrots. Am J Vet Res 2011; 72: 210–218.
14. Lee YL, Yu PH, Chen CL, et al. Determination of the enhancement effect and diameters of the major arteries of African grey parrots using a dual-head power injector for computed tomographic angiography. Taiwan Vet J 2015; 41: 165–175.
15. Hirai N, Imakita S, Tanaka R, et al. Multidetector CT angiography using a dual-head power injector and bolus tracking for the diagnosis of aortic aneurysms at reduced contrast material dosage. Acad Radiol 2006; 13: 694–700.
16. Kim DJ, Kim TH, Kim SJ, et al. Saline flush effect for enhancement of aorta and coronary arteries at multidetector CT coronary angiography. Radiology 2008; 246: 110–115.
17. Hopper KD, Mosher TJ, Kasales CJ, et al. Thoracic spiral CT: delivery of contrast material pushed with injectable saline solution in a power injector. Radiology 1997; 205: 269–271.
18. Bae KT. Intravenous contrast medium administration and scan timing at CT: considerations and approaches. Radiology 2010; 256: 32–61.
19. Johnson-Delaney CA. Practical avian cardiology. Exotic DVM 2006; 8: 78–85.
20. Strunk A, Wilson GH. Avian cardiology. Vet Clin North Am Exot Anim Pract 2003; 6: 1–28.
21. Nap AMP, Lumeij TJ, Stokhof AA. Electrocardiogram of the African grey (Psittacus erithacus) and Amazon (Amazona spp.) parrot. Avian Pathol 1992; 21: 45–53.
22. Pees M, Straub J, Krautwald-Junghanns ME. Echocardiographic examinations of 60 African grey parrots and 30 other psittacine birds. Vet Rec 2004; 155: 73–76.
23. Capitelli R, Crosta L. Overview of psittacine blood analysis and comparative retrospective study of clinical diagnosis, hematology and blood chemistry in selected psittacine species. Vet Clin North Am Exot Anim Pract 2013; 16: 71–120.
24. Pilny AA, Quesenberry KE, Bartick-Sedrish TE, et al. Evaluation of Chlamydophila psittaci infection and other risk factors for atherosclerosis in pet psittacine birds. J Am Vet Med Assoc 2012; 240: 1474–1480.
25. Ritchie BW, Harrison GJ, Harrison LR. Cardiology. In: Lumeij JT, Ritchie BW, eds. Avian medicine: principles and application. Lake Worth, Fla: Wingers Publishing Inc, 1994;695–722.
26. Dodge JT Jr, Brown BG, Bolson EL, et al. Lumen diameter of normal human coronary arteries influence of age, sex, anatomic variation, and left ventricular hypertrophy or dilation. Circulation 1992; 86: 232–246.
27. Johnson PT, Pannu HK, Fishman EK. IV contrast infusion for coronary artery CT angiography: literature review and results of a nationwide survey. AJR Am J Roentgenol 2009;192:W214–W221.
28. Lumsden JH, Mullen K. On establishing reference values. Can J Comp Med 1978; 42: 293–301.
29. Beaufrère H, Pariaut R, Rodriguez D, et al. Comparison of transcoelomic, contrast transcoelomic, and transesophageal echocardiography in anesthetized red-tailed hawks (Buteo jamaicensis). Am J Vet Res 2012; 73: 1560–1568.
30. Fleischmann D. Contrast medium administration in computed tomographic angiography. In: Rubin GD, Rofsky NM, eds. CT and MR angiography: comprehensive vascular assessment. Philadelphia: Lippincott Williams & Wilkins, 2009:129–153.
31. Slama M, Susic D, Varagic J, et al. Echocardiographic measurement of cardiac output in rats. Am J Physiol Heart Circ Physiol 2003; 284: H691–H697.
32. Butler PJ, West NH, Jones DR. Respiratory and cardiovascular responses of the pigeon to sustained, level flight in a wind-tunnel. J Exp Biol 1977; 71: 7–26.
33. Hof RP, Hof A, Sturm RP. The Doppler method for measuring cardiac output in conscious rabbits: validation studies, uses, and limitations. J Pharmacol Methods 1990; 24: 263–276.
34. Zhuang ZW, Gao L, Murakami M, et al. Arteriogenesis: noninvasive quantification with multi–detector row CT angiography and three-dimensional volume rendering in rodents. Radiology 2006; 240: 698–707.
35. Hur J, Kim YJ, Shim HS, et al. Assessment of atherosclerotic plaques in a rabbit model by delayed-phase contrast-enhanced CT angiography: comparison with histopathology. Int J Cardiovasc Imaging 2012; 28: 353–363.
36. Krautwald-Junghanns ME, Schloemer J, Pees M. Iodine-based contrast media in avian medicine. J Exot Pet Med 2008; 17: 189–197.
37. Krautwald-Junghanns ME, Schrooff S, Bartels T. Contrast studies of the cardiovascular system (angiocardiography). In: Krautwald-Junghanns ME, Pees M, Reese S, et al, eds. Diagnostic imaging of exotic pets. Hannover, Germany: Schlütersche Verlagsgesellschaft mbH & Co, 2011;32.
38. Eller A, Wuest W, Kramer M, et al. Carotid CTA: radiation exposure and image quality with the use of attenuation-based, automated kilovolt selection. AJNR Am J Neuroradiol 2014; 35: 237–241.
39. Yu L, Liu X, Leng S, et al. Radiation dose reduction in computed tomography: techniques and future perspective. Imaging Med 2009; 1: 65–84.
40. Liu Y, Hopper KD, Mauger DT, et al. CT angiography measurement of the carotid artery: optimizing visualization by manipulating window and level settings and contrast material attenuation. Radiology 2000; 217: 494–500.
41. Suzuki S, Furui S, Kaminaga T, et al. Measurement of vascular diameter in vitro by automated software for CT angiography: effects of inner diameter, density of contrast medium, and convolution kernel. AJR Am J Roentgenol 2004; 182: 1313–1317.
42. Lell MM, Anders K, Uder M, et al. New techniques in CT angiography. Radiographics 2006;26:S45–S62.
43. Bae KT, Heiken JP, Brink JA. Aortic and hepatic contrast medium enhancement at CT. I. Prediction with a computer model. Radiology 1998; 207: 647–655.
44. Tsai IC, Lee T, Chen MC, et al. Visualization of neonatal coronary arteries on multidetector row CT: ECG-gated versus non-ECG-gated technique. Pediatr Radiol 2007; 37: 818–825.
45. Jadhav SP, Golriz F, Atweh LA, et al. CT angiography of neonates and infants: comparison of radiation dose and image quality of target mode prospectively ECG-gated 320-MDCT and ungated helical 64-MDCT. AJR Am J Roentgenol 2015;204:W184–W191.
46. Paul JF. Multislice CT of the heart and great vessels in congenital heart disease patients. In: Schoepf UF, ed. CT of the heart: principles and applications. Totowa, NJ: Humana Press, 2005;161–169.
